NATURAL SCIENCES OP PHILADELPHIA. 161 



clue to any changes of color. This theory is most singularly complicated and 

 unfortunate. It surrounds the sun, said to be dark, with an apparatus con- 

 sisting of five distinct atmospheric envelopes, all regularly arranged one above 

 the other : first, a transparent envelope touching the opaque body of the sun ; 

 secondly, a fiery luminous envelope; thirdly, another transparent envelope ; 

 fourthly, another fiery luminous envelope ; fifthly, a transparent envelope sur- 

 rounding all the others. Among such a number of imaginary things, there 

 seems to be no room to imagine how changes of color could occur. Hence 

 the mention of a change of color in a star has been regarded as anomalous, as 

 an inconvenient fact, having no relation to any popular theories, and no appro- 

 priate place in the ordinary systems. Hence observations on the colors, and 

 on the changes of colors, have not been stimulated, but rather repressed by this 

 complex theory of the sun. 



Another cause for the delay in this department of Astronomy, is the difficulty 

 of deciding on the real colors of the stars. The reason why I did not myself 

 first notice the greenness of Vega, was because I had been accustomed to re- 

 gard it as blue. I relied unknowingly more on my belief than on my vision. 

 This is the same as when in twilight, or less often in broad day, we think we 

 see an object very distinctly, and on a more careful view it turns out to be 

 really something totally different in all its parts. We s'ee partly with our 

 judgment, and partly with our eyes, and it often happens that our judgments 

 warp and change the impressions on the eye. The discoverer'of the change of 

 Sirius from its former white, had been so long accustomed to regard that star 

 as of a purple or violet blue, that it was. some time after I had said it was 

 green, before he convinced himself of its green color. Then as we all had the 

 vague idea, though entirely baseless, that a great star millions of miles in cir- 

 cumference could not change so soon, he thought he must have been mistaken 

 about the violet color. And so;did I; but since my recent observations of 

 Vega, I do not think so. Sirius and Vega seem both alike to have changed 

 quite recently and suddenly from blue to green. Hundreds of observers 

 had seen Sirius through a telescope, and yet Clark, of Boston, was the 

 first to notice that it had a companion, although that companion had been 

 plainly enough in the field of view of all their telescopes. Since then, 

 Goldschmidt, of Paris, has announced, that with a far inferior instrument, 

 he has observed five more companions of Sirius, all shining in its near 

 vicinity. Previous observers did not see them, although they must have 

 received the impressions on their retinas. Direct attention was required to 

 those special objects. In a clear night we seem to see, by an optical illusion, 

 ten thousand stars. The whole heavens swarm with them, and all, on account 

 of their minuteness, appear to the naked eye to shine with a white light. The 

 milky way deepens this general impression of whiteness. Probably less than 

 fifty stars on any night, at once, are large enough to give the impression of 

 colors to the naked eye. Thus the great mass appearing white, we assume 

 that all are white, and by this means, the judgment being wrong, the colors 

 strike the retina, but are not noticed. 



While the telescope is necessary to distinguish the colors of the smaller stars, 

 I have a suspicion that the naked eye is best for stars of the first magnitude, 

 and perhaps for the second. These can be seen sufficiently well by the un- 

 aided eye, and no delicate tints are absorbed or added, as may possibly be done 

 by the glasses and specula of instruments. The disturbing effects of the at- 

 mosphere, of moonlight, and of artificial lights, may be avoided by repeating 

 the observations at different times. 



In connection with the fact that all the stars of the first magnitude visible in 

 this latitude, and at least 3 of the 6 not here visible, are colored, it may be 

 well to state my opinion, that colored stars of all magnitudes, are far more nu- 

 merous than they are generally supposed, even by Astronomers. Of 600 of the 

 brighter double stars in the great catalogue of Struve, published in 1837, more 

 1863.] 



